Evaluation of 3D Counterfactual Brain MRI Generation

2025-08-04

AI Summary

This study addresses key challenges in medical image counterfactual generation—namely, anatomical implausibility, lack of causal controllability, and inconsistent evaluation—by proposing the first causality-guided, anatomy-aware 3D MRI generation framework based on causal graphs. It models regional brain volumes as causal variables to encode anatomical dependencies, enabling controllable volumetric editing of Alzheimer’s disease–associated brain regions. A comprehensive, multi-dimensional benchmark is established on T1-weighted MRI data, systematically evaluating six generative models across compositionality, reversibility, realism, and cross-dataset generalizability. Experiments demonstrate significant improvements in physiological plausibility and causal interpretability of generated images; however, stability in non-target regions remains suboptimal. This work introduces a novel paradigm for trustworthy medical image counterfactual modeling, accompanied by a standardized evaluation protocol and empirical validation.

Technology Category

• Machine Learning: Causal Learning
• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Reasoning under Uncertainty: Causality

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications

Abstract

Counterfactual generation offers a principled framework for simulating hypothetical changes in medical imaging, with potential applications in understanding disease mechanisms and generating physiologically plausible data. However, generating realistic structural 3D brain MRIs that respect anatomical and causal constraints remains challenging due to data scarcity, structural complexity, and the lack of standardized evaluation protocols. In this work, we convert six generative models into 3D counterfactual approaches by incorporating an anatomy-guided framework based on a causal graph, in which regional brain volumes serve as direct conditioning inputs. Each model is evaluated with respect to composition, reversibility, realism, effectiveness and minimality on T1-weighted brain MRIs (T1w MRIs) from the Alzheimer's Disease Neuroimaging Initiative (ADNI). In addition, we test the generalizability of each model with respect to T1w MRIs of the National Consortium on Alcohol and Neurodevelopment in Adolescence (NCANDA). Our results indicate that anatomically grounded conditioning successfully modifies the targeted anatomical regions; however, it exhibits limitations in preserving non-targeted structures. Beyond laying the groundwork for more interpretable and clinically relevant generative modeling of brain MRIs, this benchmark highlights the need for novel architectures that more accurately capture anatomical interdependencies.

Problem

• Generating realistic 3D counterfactual brain MRIs with anatomical constraints
• Evaluating generative models for disease mechanism understanding and data generation
• Addressing challenges in preserving non-targeted brain structures during generation

Innovation

• Incorporates anatomy-guided causal graph framework
• Evaluates six generative models on multiple criteria
• Tests generalizability across different MRI datasets